In [2]:
# Suppressing warnings:
def warn(*args, **kwargs):
    pass
import warnings
warnings.warn = warn

#import necessary libraries
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.linear_model import LogisticRegression, LinearRegression
from sklearn.impute import SimpleImputer
from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import GridSearchCV, train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn.preprocessing import StandardScaler, LabelEncoder
from sklearn import svm
from sklearn.svm import SVC
from sklearn.svm import SVR
from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor, VotingClassifier, VotingRegressor
from sklearn.metrics import f1_score, mean_squared_error, confusion_matrix, accuracy_score, classification_report
from sklearn import preprocessing, metrics
import plotly.express as px
from plotly.subplots import make_subplots
import plotly.graph_objects as go
In [3]:
# Load the dataset
df = pd.read_csv('WEATHER.csv')

# Display the first few rows of the dataframe
print(df.head())

        Date  Temp9am  Temp3pm  MinTemp  MaxTemp  Rainfall RainToday  \
0  1/17/2015     20.7     20.9     19.5     22.4      15.6       Yes   
1  1/18/2015     22.4     24.8     19.5     25.6       6.0       Yes   
2  1/19/2015     23.5     23.0     21.6     24.5       6.6       Yes   
3  1/20/2015     21.4     20.9     20.2     22.8      18.8       Yes   
4  1/21/2015     22.5     25.5     19.7     25.7      77.4       Yes   

   Evaporation  Sunshine WindGustDir  ...  WindDir9am WindDir3pm WindSpeed9am  \
0          6.2       0.0           W  ...           S        SSW           17   
1          3.4       2.7           W  ...           W          E            9   
2          2.4       0.1           W  ...         ESE        ESE           17   
3          2.2       0.0           W  ...         NNE          E           22   
4          4.8       0.0           W  ...         NNE          W           11   

   WindSpeed3pm  Humidity9am  Humidity3pm  Pressure9am  Pressure3pm  Cloud9am  \
0            20           92           84       1017.6       1017.4         8   
1            13           83           73       1017.9       1016.4         7   
2             2           88           86       1016.7       1015.6         7   
3            20           83           90       1014.2       1011.8         8   
4             6           88           74       1008.3       1004.8         8   

   Cloud3pm  
0         8  
1         7  
2         8  
3         8  
4         8  

[5 rows x 21 columns]
In [4]:
#DATA Preprocessing
#Data Cleaning
# Handle Missing Values
# Let's fill missing numerical values with the mean of their respective columns
numerical_cols = df.select_dtypes(include=['float64', 'int64']).columns
df[numerical_cols] = df[numerical_cols].fillna(df[numerical_cols].mean())
In [5]:
# For categorical variables, let's fill missing values with the most frequent value
categorical_cols = df.select_dtypes(include=['object']).columns
df[categorical_cols] = df[categorical_cols].fillna(df[categorical_cols].mode().iloc[0])
In [6]:
# Convert Categorical Variables into Numerical using One-Hot Encoding for categorical variables
one_hot_encoded_df = pd.get_dummies(df, columns=categorical_cols)
In [7]:
# Separate Features and Target Variable
X = df.drop(['Date', 'RainToday', 'Rainfall'], axis=1)  # Features
y = df['Rainfall']  # Target variable
In [8]:
#Data Analysis: feature distribution
#MIN TEMP
sns.displot(df, x="MinTemp", hue='RainToday', kde=True)
plt.title("Minimum Temperature Distribution", fontsize = 14)
plt.show()
In [9]:
#Data Analysis: feature distribution
#MAX TEMP
sns.displot(df, x="MaxTemp", hue='RainToday', kde=True)
plt.title("Maximum Temperature Distribution", fontsize = 14)
plt.show()
In [10]:
#Data Analysis: feature distribution
#WINDGUSTDIST
sns.displot(df, x="WindGustSpeed", hue='RainToday', kde=True)
plt.title("Wind Gust Distribution", fontsize = 14)
plt.show()
In [11]:
#Data Analysis: feature distribution
#WINDGUSTDIST 9AM
sns.displot(df, x="WindSpeed9am", hue='RainToday', kde=True)
plt.title("WindSpeed at 9am Distribution", fontsize = 14)
plt.show()
In [12]:
#Data Analysis: feature distribution
#WINDGUSTDIST 3PM
sns.displot(df, x="WindSpeed3pm", hue='RainToday', kde=True)
plt.title("WindSpeed at 3pm Distribution", fontsize = 14)
plt.show()
In [13]:
#Data Analysis: feature distribution
#HUMIDITY 9AM
sns.displot(df, x="Humidity9am", hue='RainToday', kde=True)
plt.title("Humidity at 9am Distribution", fontsize = 14)
plt.show()
In [14]:
#Data Analysis: feature distribution
#HUMIDITY 3PM
sns.displot(df, x="Humidity3pm", hue='RainToday', kde=True)
plt.title("Humidity at 3pm Distribution", fontsize = 14)
plt.show()
In [15]:
#Data Analysis: feature distribution
#PRESSURE 9AM
sns.displot(df, x="Pressure9am", hue='RainToday', kde=True)
plt.title("Pressure at 9am Distribution", fontsize = 14)
plt.show()
In [16]:
#Data Analysis: feature distribution
#PRESSURE 3PM
sns.displot(df, x="Pressure3pm", hue='RainToday', kde=True)
plt.title("Pressure at 3pm Distribution", fontsize = 14)
plt.show()
In [17]:
#Data Analysis: feature distribution
#CLOUD-DIST 9AM
sns.displot(df, x="Cloud9am", hue='RainToday', kde=True)
plt.title("Cloud at 9am Distribution", fontsize = 14)
plt.show()
In [18]:
#Data Analysis: feature distribution
#CLOUD-DIST 3PM
sns.displot(df, x="Cloud3pm", hue='RainToday', kde=True)
plt.title("Cloud at 3pm Distribution", fontsize = 14)
plt.show()
In [19]:
#Data Analysis: feature distribution
#CLOUD-DIST 9AM
sns.displot(df, x="Temp9am", hue='RainToday', kde=True)
plt.title("Temperature at 9am Distribution", fontsize = 14)
plt.show()
In [20]:
#Data Analysis: feature distribution
#CLOUD-DIST 3PM
sns.displot(df, x="Temp3pm", hue='RainToday', kde=True)
plt.title("Temperature at 3pm Distribution", fontsize = 14)
plt.show()
In [21]:
#Data Analysis: EDA
#AVG WINDSPEED ANALYSIS

# Convert 'Date' column to datetime, handling different formats
df['Date'] = pd.to_datetime(df['Date'], errors='coerce')

# Drop rows with invalid dates
df = df.dropna(subset=['Date'])

# Extract the month from the 'Date' column
df['Month'] = df['Date'].dt.month

# Group by the extracted month and calculate mean wind speeds for 9am and 3pm
monthly_windspeed = df.groupby('Month')[['WindSpeed9am', 'WindSpeed3pm']].mean().reset_index()
In [22]:
#Data Analysis: EDA
#AVG WINDSPEED OBSERVATION (LINE PLOT) 
# Plotting
x = monthly_windspeed['Month']
y1 = monthly_windspeed['WindSpeed9am']
y2 = monthly_windspeed['WindSpeed3pm']

plt.figure(figsize=(15, 8))
plt.plot(x, y1, marker='D', color='darkgreen', label='WindSpeed at 9am')
plt.plot(x, y2, marker='D', color='darkorange', label='WindSpeed at 3pm')
plt.xlabel('Month', fontsize=14)
plt.ylabel('WindSpeed', fontsize=14)
plt.title('Month-wise observation of Average WindSpeed', fontsize=18)
plt.xticks(range(1, 13))  # Set the x-axis to only show months 1 to 12
plt.legend(fontsize=10, loc='best')
plt.grid(True)
plt.show()
In [23]:
#Data Analysis: EDA
#AVG HUMIDITY ANALYSIS

# Convert 'Date' column to datetime, handling different formats
df['Date'] = pd.to_datetime(df['Date'], errors='coerce')

# Drop rows with invalid dates
df = df.dropna(subset=['Date'])

# Extract the month from the 'Date' column
df['Month'] = df['Date'].dt.month

# Group by the extracted month and calculate mean humidity for 9am and 3pm
humidity_weather_df = df.groupby('Month')[['Humidity9am', 'Humidity3pm']].mean().reset_index()
In [24]:
#Data Analysis: EDA
#AVG HUMIDITY ANALYSIS OBSERVATION (STACKED BAR PLOT)
# Plotting
x = humidity_weather_df['Month']
y1 = humidity_weather_df['Humidity9am']
y2 = humidity_weather_df['Humidity3pm']

plt.figure(figsize=(15, 8))
bar_width = 0.35
index = x

plt.bar(index - bar_width/2, y1, bar_width, color='gold', label='Humidity at 9am')
plt.bar(index + bar_width/2, y2, bar_width, color='coral', label='Humidity at 3pm')

plt.xlabel('Month', fontsize=14)
plt.ylabel('Humidity', fontsize=14)
plt.title('Month-wise observation of Average Humidity', fontsize=18)
plt.xticks(index, range(1, 13))
plt.legend(fontsize=10, loc='best')
plt.grid(True)
plt.show()
In [25]:
#Data Analysis: EDA
#AVG PRESSURE ANALYSIS

# Convert 'Date' column to datetime, handling different formats
df['Date'] = pd.to_datetime(df['Date'], errors='coerce')

# Drop rows with invalid dates
df = df.dropna(subset=['Date'])

# Extract the month from the 'Date' column
df['Month'] = df['Date'].dt.month

# Group by the extracted month and calculate mean pressure for 9am and 3pm
pressure_weather_df = df.groupby('Month')[['Pressure9am', 'Pressure3pm']].mean().reset_index()
In [26]:
#Data Analysis: EDA
#AVG PRESSURE ANALYSIS (LINE PLOT)
# Plotting
x = pressure_weather_df['Month']
y1 = pressure_weather_df['Pressure9am']
y2 = pressure_weather_df['Pressure3pm']

plt.figure(figsize=(15, 8))
plt.plot(x, y1, marker='o', color='purple', label='Pressure at 9am')
plt.plot(x, y2, marker='o', color='darkcyan', label='Pressure at 3pm')
plt.xlabel('Month', fontsize=14)
plt.ylabel('Pressure', fontsize=14)
plt.title('Month-wise observation of Average Pressure', fontsize=18)
plt.xticks(range(1, 13))  # Set the x-axis to only show months 1 to 12
plt.legend(fontsize=10, loc='best')
plt.grid(True)
plt.show()
In [27]:
#Data Analysis: EDA
#Correlation and Description of the data
bf=df.drop(columns=['Date','RainToday','WindGustDir', 'WindDir9am', 'WindDir3pm'])
bf.corr()
Out[27]:
Temp9am Temp3pm MinTemp MaxTemp Rainfall Evaporation Sunshine WindGustSpeed WindSpeed9am WindSpeed3pm Humidity9am Humidity3pm Pressure9am Pressure3pm Cloud9am Cloud3pm Month
Temp9am 1.000000 0.828457 0.939307 0.858639 -0.054297 0.606122 0.113891 0.159570 -0.334084 0.280009 -0.145944 0.131855 -0.417437 -0.401353 0.093204 0.068683 0.269298
Temp3pm 0.828457 1.000000 0.753147 0.960805 -0.142748 0.474599 0.345528 0.036216 -0.351836 0.136713 -0.155192 -0.212696 -0.342862 -0.387945 -0.086544 -0.144965 0.223978
MinTemp 0.939307 0.753147 1.000000 0.771005 0.013995 0.569316 -0.063475 0.127792 -0.298040 0.225905 0.053946 0.273269 -0.391654 -0.353334 0.241248 0.164028 0.284149
MaxTemp 0.858639 0.960805 0.771005 1.000000 -0.139306 0.510334 0.327422 0.073184 -0.349382 0.155542 -0.185382 -0.150741 -0.385853 -0.420375 -0.084695 -0.099217 0.229074
Rainfall -0.054297 -0.142748 0.013995 -0.139306 1.000000 -0.110343 -0.308962 0.149811 0.153696 0.039415 0.332636 0.305635 -0.037561 0.001780 0.245231 0.200999 0.007841
Evaporation 0.606122 0.474599 0.569316 0.510334 -0.110343 1.000000 0.175932 0.245426 -0.077532 0.292469 -0.395651 -0.102295 -0.343581 -0.298303 -0.037051 -0.038422 0.161738
Sunshine 0.113891 0.345528 -0.063475 0.327422 -0.308962 0.175932 1.000000 -0.029032 -0.058410 0.177813 -0.490631 -0.586476 -0.047789 -0.094507 -0.665014 -0.677861 -0.010061
WindGustSpeed 0.159570 0.036216 0.127792 0.073184 0.149811 0.245426 -0.029032 1.000000 0.256460 0.509825 -0.225613 -0.095037 -0.408876 -0.351850 0.024275 0.044016 -0.021639
WindSpeed9am -0.334084 -0.351836 -0.298040 -0.349382 0.153696 -0.077532 -0.058410 0.256460 1.000000 0.197669 -0.189763 -0.148849 0.003737 0.091878 -0.062441 -0.018030 -0.080324
WindSpeed3pm 0.280009 0.136713 0.225905 0.155542 0.039415 0.292469 0.177813 0.509825 0.197669 1.000000 -0.288910 -0.135039 -0.333810 -0.277749 -0.050932 -0.095435 0.059054
Humidity9am -0.145944 -0.155192 0.053946 -0.185382 0.332636 -0.395651 -0.490631 -0.225613 -0.189763 -0.288910 1.000000 0.657933 0.262098 0.249202 0.431550 0.331062 0.017233
Humidity3pm 0.131855 -0.212696 0.273269 -0.150741 0.305635 -0.102295 -0.586476 -0.095037 -0.148849 -0.135039 0.657933 1.000000 0.151608 0.179611 0.463529 0.472248 0.081332
Pressure9am -0.417437 -0.342862 -0.391654 -0.385853 -0.037561 -0.343581 -0.047789 -0.408876 0.003737 -0.333810 0.262098 0.151608 1.000000 0.963307 -0.015889 -0.039509 -0.099024
Pressure3pm -0.401353 -0.387945 -0.353334 -0.420375 0.001780 -0.298303 -0.094507 -0.351850 0.091878 -0.277749 0.249202 0.179611 0.963307 1.000000 0.013787 -0.003536 -0.090527
Cloud9am 0.093204 -0.086544 0.241248 -0.084695 0.245231 -0.037051 -0.665014 0.024275 -0.062441 -0.050932 0.431550 0.463529 -0.015889 0.013787 1.000000 0.604359 0.034927
Cloud3pm 0.068683 -0.144965 0.164028 -0.099217 0.200999 -0.038422 -0.677861 0.044016 -0.018030 -0.095435 0.331062 0.472248 -0.039509 -0.003536 0.604359 1.000000 0.051716
Month 0.269298 0.223978 0.284149 0.229074 0.007841 0.161738 -0.010061 -0.021639 -0.080324 0.059054 0.017233 0.081332 -0.099024 -0.090527 0.034927 0.051716 1.000000
In [28]:
df.describe()
Out[28]:
Date Temp9am Temp3pm MinTemp MaxTemp Rainfall Evaporation Sunshine WindGustSpeed WindSpeed9am WindSpeed3pm Humidity9am Humidity3pm Pressure9am Pressure3pm Cloud9am Cloud3pm Month
count 3271 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000 3271.000000
mean 2019-07-10 00:00:00 17.821461 21.543656 14.877102 23.005564 3.342158 5.175787 7.168970 41.476307 15.077041 19.294405 68.243962 54.698563 1018.334424 1016.003085 4.318557 4.176093 6.550290
min 2015-01-17 00:00:00 6.400000 10.200000 4.300000 11.700000 0.000000 0.000000 0.000000 17.000000 0.000000 0.000000 19.000000 10.000000 986.700000 989.800000 0.000000 0.000000 1.000000
25% 2017-04-13 12:00:00 13.800000 18.400000 11.000000 19.600000 0.000000 3.200000 4.250000 35.000000 11.000000 15.000000 58.000000 44.000000 1013.700000 1011.300000 2.000000 2.000000 4.000000
50% 2019-07-10 00:00:00 18.200000 21.300000 14.900000 22.800000 0.000000 4.800000 8.300000 41.000000 15.000000 19.000000 69.000000 56.000000 1018.600000 1016.300000 5.000000 4.000000 7.000000
75% 2021-10-04 12:00:00 21.700000 24.500000 18.800000 26.000000 1.400000 7.000000 10.200000 44.000000 20.000000 24.000000 80.000000 64.000000 1023.100000 1020.800000 7.000000 7.000000 10.000000
max 2023-12-31 00:00:00 36.500000 44.700000 27.600000 45.800000 119.400000 18.400000 13.600000 96.000000 54.000000 57.000000 100.000000 99.000000 1039.000000 1036.700000 9.000000 8.000000 12.000000
std NaN 4.894316 4.297053 4.554710 4.483752 9.917746 2.757684 3.815966 10.806951 7.043825 7.453331 15.086127 16.279241 7.020090 7.019915 2.526923 2.411274 3.435802
In [29]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3271 entries, 0 to 3270
Data columns (total 22 columns):
 #   Column         Non-Null Count  Dtype         
---  ------         --------------  -----         
 0   Date           3271 non-null   datetime64[ns]
 1   Temp9am        3271 non-null   float64       
 2   Temp3pm        3271 non-null   float64       
 3   MinTemp        3271 non-null   float64       
 4   MaxTemp        3271 non-null   float64       
 5   Rainfall       3271 non-null   float64       
 6   RainToday      3271 non-null   object        
 7   Evaporation    3271 non-null   float64       
 8   Sunshine       3271 non-null   float64       
 9   WindGustDir    3271 non-null   object        
 10  WindGustSpeed  3271 non-null   int64         
 11  WindDir9am     3271 non-null   object        
 12  WindDir3pm     3271 non-null   object        
 13  WindSpeed9am   3271 non-null   int64         
 14  WindSpeed3pm   3271 non-null   int64         
 15  Humidity9am    3271 non-null   int64         
 16  Humidity3pm    3271 non-null   int64         
 17  Pressure9am    3271 non-null   float64       
 18  Pressure3pm    3271 non-null   float64       
 19  Cloud9am       3271 non-null   int64         
 20  Cloud3pm       3271 non-null   int64         
 21  Month          3271 non-null   int32         
dtypes: datetime64[ns](1), float64(9), int32(1), int64(7), object(4)
memory usage: 549.6+ KB
In [30]:
#Data Analysis: EDA
plt.figure(figsize=(20, 20))
sns.heatmap(bf.corr(), annot=True)
Out[30]:
<Axes: >
In [31]:
# Selecting the columns to be scaled
bf = df[['Temp9am', 'Temp3pm', 'MinTemp', 'MaxTemp', 'Rainfall', 
         'Evaporation', 'Sunshine', 'WindGustSpeed', 'WindSpeed9am', 
         'WindSpeed3pm', 'Humidity9am', 'Humidity3pm', 'Pressure9am', 
         'Pressure3pm', 'Cloud9am', 'Cloud3pm']]

#Standardize the Variables
# Scaling the features
scaler = StandardScaler()
scaled_features = scaler.fit_transform(bf)
df_feat = pd.DataFrame(scaled_features, columns=bf.columns)
df_feat = pd.DataFrame(scaled_features,columns=bf.columns[:16])
# Adding back the non-scaled columns
selected_columns = df[['Date','RainToday','WindGustDir', 'WindDir9am', 'WindDir3pm']]
df_feat = pd.concat([df_feat, selected_columns.reset_index(drop=True)], axis=1)
df_feat.head()
Out[31]:
Temp9am Temp3pm MinTemp MaxTemp Rainfall Evaporation Sunshine WindGustSpeed WindSpeed9am WindSpeed3pm ... Humidity3pm Pressure9am Pressure3pm Cloud9am Cloud3pm Date RainToday WindGustDir WindDir9am WindDir3pm
0 0.588229 -0.149813 1.015126 -0.135078 1.236139 0.371460 -1.878965 -0.044081 0.273041 0.094683 ... 1.800202 -0.104633 0.199024 1.457110 1.586088 2015-01-17 Yes W S SSW
1 0.935624 0.757924 1.015126 0.578719 0.268029 -0.644040 -1.171303 -0.044081 -0.862879 -0.844638 ... 1.124391 -0.061892 0.056550 1.061312 1.171306 2015-01-18 Yes W W E
2 1.160409 0.338969 1.476258 0.333351 0.328536 -1.006719 -1.852755 -0.044081 0.273041 -2.320714 ... 1.923076 -0.232857 -0.057429 1.061312 1.586088 2015-01-19 Yes W ESE ESE
3 0.731274 -0.149813 1.168837 -0.045853 1.558843 -1.079254 -1.878965 -0.044081 0.982991 0.094683 ... 2.168826 -0.589032 -0.598829 1.457110 1.586088 2015-01-20 Yes W NNE E
4 0.956059 0.920852 1.059043 0.601025 7.468346 -0.136290 -1.878965 -0.044081 -0.578899 -1.783959 ... 1.185828 -1.429605 -1.596144 1.457110 1.586088 2015-01-21 Yes W NNE W

5 rows × 21 columns

In [32]:
#Transforming Categorical Variables:to convert categorical variables to binary variables to select what only neeed
# One-hot encoding for categorical variables
df_proc = pd.get_dummies(df_feat, columns=['WindGustDir', 'WindDir9am', 'WindDir3pm'])
df_proc.head()
Out[32]:
Temp9am Temp3pm MinTemp MaxTemp Rainfall Evaporation Sunshine WindGustSpeed WindSpeed9am WindSpeed3pm ... WindDir3pm_NNW WindDir3pm_NW WindDir3pm_S WindDir3pm_SE WindDir3pm_SSE WindDir3pm_SSW WindDir3pm_SW WindDir3pm_W WindDir3pm_WNW WindDir3pm_WSW
0 0.588229 -0.149813 1.015126 -0.135078 1.236139 0.371460 -1.878965 -0.044081 0.273041 0.094683 ... False False False False False True False False False False
1 0.935624 0.757924 1.015126 0.578719 0.268029 -0.644040 -1.171303 -0.044081 -0.862879 -0.844638 ... False False False False False False False False False False
2 1.160409 0.338969 1.476258 0.333351 0.328536 -1.006719 -1.852755 -0.044081 0.273041 -2.320714 ... False False False False False False False False False False
3 0.731274 -0.149813 1.168837 -0.045853 1.558843 -1.079254 -1.878965 -0.044081 0.982991 0.094683 ... False False False False False False False False False False
4 0.956059 0.920852 1.059043 0.601025 7.468346 -0.136290 -1.878965 -0.044081 -0.578899 -1.783959 ... False False False False False False False True False False

5 rows × 66 columns

In [33]:
# replace the values of the 'RainToday' column changing them from a categorical column to a binary column. 
# Replace 'Yes'/'No' with 1/0 in 'RainToday' column
df_proc['RainToday'].replace(['No', 'Yes'], [0, 1], inplace=True)
df_proc[['Date','RainToday']]
Out[33]:
Date RainToday
0 2015-01-17 1
1 2015-01-18 1
2 2015-01-19 1
3 2015-01-20 1
4 2015-01-21 1
... ... ...
3266 2023-12-27 0
3267 2023-12-28 0
3268 2023-12-29 0
3269 2023-12-30 0
3270 2023-12-31 0

3271 rows × 2 columns

In [34]:
# Drop the 'Date' column
df_proc.drop('Date', axis=1, inplace=True)
In [35]:
#Data Preprocessing
#Rainfall Occurance Prediction

# Convert all data to float type
df_proc = df_proc.astype(float)

# Defining features and target variable
features = df_proc.drop(columns='RainToday')
Y = df_proc['RainToday']

# Splitting the data into training and testing sets
x_train, x_test, y_train, y_test = train_test_split(features, Y, test_size=0.35, random_state=101)

# Display the first few rows of the processed dataframe and the shapes of train and test sets
df_proc.head(), x_train.shape, x_test.shape, y_train.shape, y_test.shape
Out[35]:
(    Temp9am   Temp3pm   MinTemp   MaxTemp  Rainfall  Evaporation  Sunshine  \
 0  0.588229 -0.149813  1.015126 -0.135078  1.236139     0.371460 -1.878965   
 1  0.935624  0.757924  1.015126  0.578719  0.268029    -0.644040 -1.171303   
 2  1.160409  0.338969  1.476258  0.333351  0.328536    -1.006719 -1.852755   
 3  0.731274 -0.149813  1.168837 -0.045853  1.558843    -1.079254 -1.878965   
 4  0.956059  0.920852  1.059043  0.601025  7.468346    -0.136290 -1.878965   
 
    WindGustSpeed  WindSpeed9am  WindSpeed3pm  ...  WindDir3pm_NNW  \
 0      -0.044081      0.273041      0.094683  ...             0.0   
 1      -0.044081     -0.862879     -0.844638  ...             0.0   
 2      -0.044081      0.273041     -2.320714  ...             0.0   
 3      -0.044081      0.982991      0.094683  ...             0.0   
 4      -0.044081     -0.578899     -1.783959  ...             0.0   
 
    WindDir3pm_NW  WindDir3pm_S  WindDir3pm_SE  WindDir3pm_SSE  WindDir3pm_SSW  \
 0            0.0           0.0            0.0             0.0             1.0   
 1            0.0           0.0            0.0             0.0             0.0   
 2            0.0           0.0            0.0             0.0             0.0   
 3            0.0           0.0            0.0             0.0             0.0   
 4            0.0           0.0            0.0             0.0             0.0   
 
    WindDir3pm_SW  WindDir3pm_W  WindDir3pm_WNW  WindDir3pm_WSW  
 0            0.0           0.0             0.0             0.0  
 1            0.0           0.0             0.0             0.0  
 2            0.0           0.0             0.0             0.0  
 3            0.0           0.0             0.0             0.0  
 4            0.0           1.0             0.0             0.0  
 
 [5 rows x 65 columns],
 (2126, 64),
 (1145, 64),
 (2126,),
 (1145,))
In [36]:
#Logistic Regression Model
LR = LogisticRegression()
LR.fit(x_train,y_train)
Out[36]:
LogisticRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
LogisticRegression()
In [37]:
#KNN Classification Model
KNN = KNeighborsClassifier(n_neighbors=4)
KNN.fit(x_train, y_train)
Out[37]:
KNeighborsClassifier(n_neighbors=4)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
KNeighborsClassifier(n_neighbors=4)
In [38]:
#Decision Tree Classification Model
Tree = DecisionTreeClassifier()
Tree.fit(x_train,y_train)
Out[38]:
DecisionTreeClassifier()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
DecisionTreeClassifier()
In [39]:
#SVM Classification Model
SVM = svm.SVC(kernel='linear')
SVM.fit(x_train, y_train)
Out[39]:
SVC(kernel='linear')
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
SVC(kernel='linear')
In [40]:
#Random Forest Classification Model
RFC = RandomForestClassifier(n_estimators=100)
RFC.fit(x_train, y_train)
Out[40]:
RandomForestClassifier()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
RandomForestClassifier()
In [42]:
# Convert x_test to a numpy array
x_test_array = np.array(x_test)

# Calculate accuracy scores
LR_Accuracy_Score = metrics.accuracy_score(y_test, LR.predict(x_test_array))
KNN_Accuracy_Score = metrics.accuracy_score(y_test, KNN.predict(x_test_array))
Tree_Accuracy_Score = metrics.accuracy_score(y_test, Tree.predict(x_test_array))
SVM_Accuracy_Score = metrics.accuracy_score(y_test, SVM.predict(x_test_array))
RFC_Accuracy_Score = metrics.accuracy_score(y_test, RFC.predict(x_test_array))

# Calculate F1 scores
LR_F1_Score = metrics.f1_score(y_test, LR.predict(x_test_array))
KNN_F1_Score = metrics.f1_score(y_test, KNN.predict(x_test_array), average='weighted')
Tree_F1_Score = metrics.f1_score(y_test, Tree.predict(x_test_array))
SVM_F1_Score = metrics.f1_score(y_test, SVM.predict(x_test_array))
RFC_F1_Score = metrics.f1_score(y_test, RFC.predict(x_test_array))

# Create the report
Report = {
    'Classification Algorithm': [
        'Logistic Regression', 'KNN', 'Decision Tree', 'SVM', 'Random Forest'
    ],
    'Accuracy Score': [
        LR_Accuracy_Score, KNN_Accuracy_Score, Tree_Accuracy_Score, SVM_Accuracy_Score, RFC_Accuracy_Score
    ],
    'F1-score': [
        LR_F1_Score, KNN_F1_Score, Tree_F1_Score, SVM_F1_Score, RFC_F1_Score
    ]
}

Report = pd.DataFrame(Report)
Report
Out[42]:
Classification Algorithm Accuracy Score F1-score
0 Logistic Regression 0.823581 0.620301
1 KNN 0.800000 0.774730
2 Decision Tree 0.773799 0.584270
3 SVM 0.828821 0.618677
4 Random Forest 0.829694 0.609218
In [43]:
#visualizing of accuracies 
fig = make_subplots(rows=1, cols=1)
fig.add_trace(
    go.Bar(x=Report['Classification Algorithm'], y=Report['Accuracy Score'], name='Accuracy Score'),
    row=1, col=1
)

fig.add_trace(
    go.Bar(x=Report['Classification Algorithm'], y=Report['F1-score'], name='F1-score'), row=1,col=1
)
fig.update_layout(title_text="Accuracy Scores Of Classification Algorithms")
fig.update_yaxes(title_text="Score", row=1, col=1)
fig.update_xaxes(title_text="Classification Algorithm", row=1, col=1)
fig.show()
In [44]:
#Data Preprocessing
#Rainfall Amount Prediction
df_proc = df_proc.astype(float)
features = df_proc.drop(columns='Rainfall', axis=1)
Y = df_proc['Rainfall']
x_train, x_test, y_train, y_test = train_test_split(features, Y, test_size=0.35, random_state=101)
In [45]:
#Linear Regression
LinearReg = LinearRegression()
LinearReg.fit(x_train, y_train)
Out[45]:
LinearRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
LinearRegression()
In [46]:
#Random Forest Regression Model
RFR = RandomForestRegressor()
RFR.fit(x_train, y_train)
Out[46]:
RandomForestRegressor()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
RandomForestRegressor()
In [47]:
# SVM Regression Model
SVRR = SVR()
SVRR.fit(x_train, y_train)
Out[47]:
SVR()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
SVR()
In [48]:
#Regression report on rainfall amount prediction
LinearRegression_MAE = metrics.mean_absolute_error(LinearReg.predict(x_test), y_test)
LinearRegression_MSE = metrics.mean_squared_error(LinearReg.predict(x_test), y_test, squared=False)

RandomForest_MAE = metrics.mean_absolute_error(RFR.predict(x_test), y_test)
RandomForest_MSE = metrics.mean_squared_error(RFR.predict(x_test), y_test, squared=False)

SVM_MAE = metrics.mean_absolute_error(SVRR.predict(x_test), y_test)
SVM_MSE = metrics.mean_squared_error(SVRR.predict(x_test), y_test, squared=False)

Report = {'Regression Algorithm': ['Linear Regression','Random Forest', 'SVM'],
          'MAE': [LinearRegression_MAE,RandomForest_MAE, SVM_MAE],
          'RMSE': [LinearRegression_MSE,RandomForest_MSE, SVM_MSE],}
Report = pd.DataFrame(Report)
Report
Out[48]:
Regression Algorithm MAE RMSE
0 Linear Regression 0.498764 0.948275
1 Random Forest 0.380826 0.888466
2 SVM 0.365070 0.971967
In [49]:
#visualizing of accuracies 
fig = make_subplots(rows=1, cols=1)

fig.add_trace(
    go.Bar(x=Report['Regression Algorithm'], y=Report['MAE'], name='MAE'),
    row=1, col=1
)

fig.add_trace(
    go.Bar(x=Report['Regression Algorithm'], y=Report['RMSE'], name='RMSE'), row=1,col=1
)
fig.update_layout(title_text="Error of Regression Algorithms")
fig.update_yaxes(title_text="Error", row=1, col=1)
fig.update_xaxes(title_text="Regression Algorithm", row=1, col=1)
fig.show()
In [50]:
#Average Temperature of a day Prediction
df_proc = df_proc.astype(float)
df_proc['AvgTemp']=df_proc[['Temp9am', 'Temp3pm','MinTemp','MaxTemp']].mean(axis=1)
features = df_proc.drop(columns=['Temp9am', 'Temp3pm','MinTemp','MaxTemp','AvgTemp'], axis=1)
Y = df_proc['AvgTemp']
x_train, x_test, y_train, y_test = train_test_split(features, Y, test_size=0.35, random_state=101)
In [51]:
#Linear Regression
LinearReg = LinearRegression()
LinearReg.fit(x_train, y_train)
Out[51]:
LinearRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
LinearRegression()
In [52]:
#Random Forest Regression Model
RFR = RandomForestRegressor()
RFR.fit(x_train, y_train)
Out[52]:
RandomForestRegressor()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
RandomForestRegressor()
In [53]:
#SVM Regression Model
SVR=SVR()
SVR.fit(x_train, y_train)
Out[53]:
SVR()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
SVR()
In [54]:
#Regression report on average temperature prediction
LinearRegression_MAE = metrics.mean_absolute_error(LinearReg.predict(x_test), y_test)
LinearRegression_MSE = metrics.mean_squared_error(LinearReg.predict(x_test), y_test, squared=False)

RandomForest_MAE = metrics.mean_absolute_error(RFR.predict(x_test), y_test)
RandomForest_MSE = metrics.mean_squared_error(RFR.predict(x_test), y_test, squared=False)

SVM_MAE = metrics.mean_absolute_error(SVR.predict(x_test), y_test)
SVM_MSE = metrics.mean_squared_error(SVR.predict(x_test), y_test, squared=False)

Report = {'Regression Algorithm': ['Linear Regression','Random Forest', 'SVM'],
          'MAE': [LinearRegression_MAE,RandomForest_MAE, SVM_MAE],
          'RMSE': [LinearRegression_MSE,RandomForest_MSE, SVM_MSE],}
Report = pd.DataFrame(Report)
Report
Out[54]:
Regression Algorithm MAE RMSE
0 Linear Regression 0.470596 0.603210
1 Random Forest 0.448228 0.571540
2 SVM 0.434701 0.560317
In [114]:
#visualization of error
fig = make_subplots(rows=1, cols=1)

fig.add_trace(
    go.Bar(x=Report['Regression Algorithm'], y=Report['MAE'], name='MAE'),
    row=1, col=1
)

fig.add_trace(
    go.Bar(x=Report['Regression Algorithm'], y=Report['RMSE'], name='RMSE'), row=1,col=1
)
fig.update_layout(title_text="Error of Regression Algorithms")
fig.update_yaxes(title_text="Error", row=1, col=1)
fig.update_xaxes(title_text="Regression Algorithm", row=1, col=1)
fig.show()